Electoral endorsement and candidate questionnaire on AI

Civil-society organisations issue candidate questionnaires, endorsement grades, and public voter-guide scorecards on AI-and-technology issues in a national, state, or local election cycle — modelled on the Sierra Club Voter Guide, the League of Conservation Voters Scorecard, the NRA's candidate ratings, or the Right to Life endorsement machinery. The organisation defines a set of positions (support for a bill, opposition to a rule, commitment to appointment criteria for a regulatory body) it wants candidates to declare on, sends a structured questionnaire, publishes the responses and non-responses, and issues an endorsement grade the organisation's supporters can carry into the polling booth. The output is a public candidate-response record that structures the next legislative cycle: incumbents who declined to respond or endorsed positions the coalition opposed carry that record forward into the next campaign cycle.

An actor chooses this strategy because electoral cycles are one of the few moments a legislator's public position on the movement's issue is scheduled and forced — the candidate cannot avoid the questionnaire the way they can avoid a lobby meeting, and their answer or non-answer becomes a permanent public artifact. The strategy converts a diffuse policy demand into a specific vote (yours) the movement's constituency knows to withhold or extend on the basis of. It also builds durable political infrastructure: the questionnaire archive lets the movement track a legislator's positions across multiple cycles, and the endorsement grade converts the movement's membership into a bloc a legislator has to weigh against opposing pressures.

It trades depth for breadth. A candidate questionnaire cannot ask the AI vendor's actual counterparty (the CEO, the lab-safety officer, the model deployer) anything — it reaches only the legislative and executive candidates who will vote on the regulation of that vendor, one step removed. The strategy also depends on the movement's membership being organised enough to actually vote the endorsement — a scorecard read only by policy insiders is a lobbying tool, not an electoral one. And AI as an electoral issue is still emergent: the salience needed to move a real bloc of voters on AI questions has been demonstrated on narrow slices (deepfake regulation, child safety) but not yet on the broader AI-governance frame the strategy would want to be voting on.

Ecology

Distinct from ballot initiative and referendum on AI. That strategy puts the measure directly before voters, bypassing the candidate. This strategy uses the candidate cycle as the forcing function for the candidate's declared position, which then determines their subsequent vote on the measure in the legislature. Ballot initiatives are direct democracy; endorsement scorecards are representative-democracy leverage.

Distinct from mass participation in agency rulemaking comments. Rulemaking comments target the executive / administrative layer — the agency's rulemaking record — outside the election cycle. Electoral endorsement targets the legislative / executive-candidate layer, on the schedule of the election. The two strategies run parallel: a well-organised coalition uses rulemaking comments to shape the administrative record and endorsement scorecards to shape the political incentive around ratifying that record.

A tactical companion to coalition lobbying of binding regional regulation. The lobby coalition working a specific bill uses the electoral questionnaire and endorsement cycle to convert declared candidate positions into legislative-vote pressure once the candidate is elected. The lobbying arm works the legislator; the endorsement arm works the electorate; the two combine into a two-front pressure the legislator's normal incumbency-advantage calculation cannot ignore.

Fed by counter-narrative framing. For a candidate questionnaire on AI to move votes, the electorate has to have a frame in which the candidate's answer matters — a narrative in which AI regulation is a live civic question, not a technical-abstraction the candidate can safely equivocate on. The frame-building work is what makes an endorsement scorecard actually mobilise voters rather than just document positions.